Tootfinder

Kathi asks ChatGPT "how to test a graph algo".
Repeats the query 6 months later…now it mentions PBT.
Huh! We ask it why it mentioned PBT. It cites back…
Our own papers and blog, and specifically mentions our own education research work on why graph algos are a great eg.
Moving the needle!

An Empirical Study of Testing Practices in Open Source AI Agent Frameworks and Agentic Applications
Mohammed Mehedi Hasan, Hao Li, Emad Fallahzadeh, Gopi Krishnan Rajbahadur, Bram Adams, Ahmed E. Hassan
https://arxiv.org/abs/2509.19185

Amazon is testing AR glasses for delivery drivers, using AI and computer vision to help them scan packages, follow walking directions, and get proof of delivery (Todd Bishop/GeekWire)
https://www.geekwire.com/2025/amazon-unveils-ai-powe…

The FDA Often Doesn't Test Generic Drugs for Quality Concerns, So ProPublica Did (ProPublica)
https://www.propublica.org/article/fda-generic-drug-testing
http://www.memeorandum.com/251222/p66#a251222p66

LED Lighting: Mini Reviews - Real-world testing! - https://m.earth.org.uk/LED-lighting.html

LED Lighting: Mini Reviews
Real-world testing!

For weeks earlier this year, the Army’s top uniformed lawyer had been raising legal concerns inside the Pentagon about some of the new policies being rolled out dictating how the military can be used and staffed.
In late January, Lt. Gen. Joe Berger, who had taken the top posting in July 2024, was asked for his advice about the legality of using Texas National Guard soldiers for immigration enforcement.
Berger told Army chief of staff Gen. Randy George that he was skeptical and wa…

How the Pentagon sidelined lawyers while testing the legal limits of military action
For weeks earlier this year, the Army’s top uniformed lawyer had been raising legal concerns inside the Pentagon about some of the new policies being rolled out dictating how the military can be used and staffed.

The Illusion of Readiness: Stress Testing Large Frontier Models on Multimodal Medical Benchmarks
Yu Gu, Jingjing Fu, Xiaodong Liu, Jeya Maria Jose Valanarasu, Noel Codella, Reuben Tan, Qianchu Liu, Ying Jin, Sheng Zhang, Jinyu Wang, Rui Wang, Lei Song, Guanghui Qin, Naoto Usuyama, Cliff Wong, Cheng Hao, Hohin Lee, Praneeth Sanapathi, Sarah Hilado, Bian Jiang, Javier Alvarez-Valle, Mu Wei, Jianfeng Gao, Eric Horvitz, Matt Lungren, Hoifung Poon, Paul Vozila

The Illusion of Readiness: Stress Testing Large Frontier Models on Multimodal Medical Benchmarks
Large frontier models like GPT-5 now achieve top scores on medical benchmarks. But our stress tests tell a different story. Leading systems often guess correctly even when key inputs like images are removed, flip answers under trivial prompt changes, and fabricate convincing yet flawed reasoning. These aren't glitches; they expose how today's benchmarks reward test-taking tricks over medical understanding. We evaluate six flagship models across six widely used benchmarks and find that high lead…

I read about a major update of #jellyfin and that one should do this planned with some care.
So I sat down, prepared for an update full of hassles, debugging, restoring, ...
> docker down / pull / validate / up / done
Took me about 5-10 min.
Very well done Jellyfin!
----

Jellyfin 10.11.0 | Jellyfin
We are pleased to bring you Jellyfin 10.11.0, our new stable release. This is probably one of, if not the, biggest and most impactful releases we've done yet, with some massive backend changes to improve performance and long-term expandability and maintainability. This release has been a long time coming, with over 6 months of development and another 6 months of RC testing, throwing our planned 6-month release schedule completely out of whack, but we definitely think the results are worth it - …

@…
> I don't think this is normal installation method on linux
Correct. This is an installer I wrote to allow multiple side by side installs, that we use for testing
> and it's not complaint with free desktop standard.
Actually to some extent it is, I am registering with the desktop environment, which is why it has dock …

DRISHTIKON: A Multimodal Multilingual Benchmark for Testing Language Models' Understanding on Indian Culture
Arijit Maji, Raghvendra Kumar, Akash Ghosh, Anushka, Nemil Shah, Abhilekh Borah, Vanshika Shah, Nishant Mishra, Sriparna Saha
https://arxiv.org/abs/2509.19274

DRISHTIKON: A Multimodal Multilingual Benchmark for Testing Language Models' Understanding on Indian Culture
We introduce DRISHTIKON, a first-of-its-kind multimodal and multilingual benchmark centered exclusively on Indian culture, designed to evaluate the cultural understanding of generative AI systems. Unlike existing benchmarks with a generic or global scope, DRISHTIKON offers deep, fine-grained coverage across India's diverse regions, spanning 15 languages, covering all states and union territories, and incorporating over 64,000 aligned text-image pairs. The dataset captures rich cultural themes i…

Since the last Android release @… allows using a custom server from which to download map files.
This decreases reliance on CoMaps-run infrastructure but is also great for people who live in places with limited internet connectivity, as you can now serve maps from local networks.
To make it easier, I wrote a small CLI tool over the weekend that downloads maps of interest & then serves them locally. A first alpha release for testing is on Codeberg:
#OpenStreetMap

A Sequential Testing Problem with Signal Control
Steven Campbell, Georgy Gaitsgori, Richard Groenewald
https://arxiv.org/abs/2509.18209 https://arxiv.org/p…

A Sequential Testing Problem with Signal Control
We study a controlled version of the Bayesian sequential testing problem for the drift of a Wiener process, in which the observer exercises discretion over the signal intensity. This control incurs a running cost that reflects the resource demands of information acquisition. The objective is to minimize the total expected cost, combining both the expenditure on control and the loss from misclassifying the unknown drift. By allowing for a general class of loss functions and any measurable cost o…

Experimentally Testing AI-Powered Content Transformations on Student Learning
Courtney Heldreth, Laura M. Vardoulakis, Nicole E. Miller, Yael Haramaty, Diana Akrong, Lidan Hackmon, Lior Belinsky
https://arxiv.org/abs/2509.18664

Posting this here for others with #LongCovid.
One of the things I've done this year is rule out potential other causes of my symptoms (imaging, blood work, neurological testing). This, in addition to treating symptoms as best I can (insomnia, brain fog, joint pain, double vision, tinnitus, hearing loss, etc.).
This fall, I wrapped up with neurology (I didn't want to, a…

Testing Aurora, the KDE brother of Bluefin, briefly. Switched to the latest branch of course. Interesting and well executed. Too much for me, application wise, nevertheless very nicely done.
PS: Bazaar is a very nice appstore and the Curated section allows Aurora (and thus others) to show their preferred applications. Very neat.

Just added Web Reachability API (at least that’s what I’m calling it) support to https://ip.small-web.org.
It’s for testing the reachability of your Small Web servers (using a domain or, more importantly, an IPv4/IPv6 address). I’m using it to implement Web Numbers¹ support in Auto Encrypt² and Kitten³.
…

X is testing a new way of opening links without fully covering an X post, allowing users to see the Like, Repost, and other buttons, starting on iOS (Cheyenne MacDonald/Engadget)
https://www.engadget.com/x-is-testing-a-new-way-of…

X is testing a new way of opening links in posts to improve engagement
The new link popups will allow the reaction buttons to remain visible so users can interact with the post while reading.

5GC-Bench: A Framework for Stress-Testing and Benchmarking 5G Core VNFs
Ioannis Panitsas, Tolga O. Atalay, Dragoslav Stojadinovic, Angelos Stavrou, Leandros Tassiulas
https://arxiv.org/abs/2509.18443

Testing ZonePane
https://fedibird.com/@takke/115416407802913320

たけうちひろあき :zonepane: 🐧 (@takke@fedibird.com)
ZonePane for iOS の Mastodon 対応版 Beta2 を TestFlight で公開しました。 ✓「返信」ができるようになりました！ ✓プロフィール画面の @トゥートを送信する、メッセージを送信する ができるようになりました！ https://testflight.apple.com/join/xu5EgEG8

Heute vor 74 Jahren: Am 24. September 1951 testete die #Sowjetunion ihre zweite #Atombombe, die RDS-2, am Semipalatinsk-Testgelände. Die Detonation, auch "Joe-2" genannt, hatte eine Sprengkraft von 38,3 Kilotonnen und wurde auf einem 30 Meter hohen Turm gezündet.

An updated Classic.

Testing, with the help of a funnel, the theory that cats are actually a liquid. Still having trouble getting his butt through the hole. #CatsOfMastodon

Aligned, Multiple-transient Events in the First #PalomarSkySurvey / Transients in the Palomar Observatory Sky Survey (POSS-I) may be associated with nuclear testing and reports of unidentified anomalous phenomena: https://iopscience.iop.org/article/10.1088/1538-3873/ae0afe / https://www.nature.com/articles/s41598-025-21620-3 -> Unexpected patterns in historical astronomical observations: https://www.su.se/english/news/unexpected-patterns-in-historical-astronomical-observations-1.855042 -> summary of why these papers are doubtful: https://x.com/MickWest/status/1980690132543107204 - links to deep discussions.

is this thing on? testing my new mastodon server :-)

These goofs want to use "drones that would fly up to 150 feet and eventually deliver some of the orders that today are transported by the company’s drivers."
We have a technology that efficiently delivers goods to homes in an urban neighborhood. It's called bicycles. https://

DoorDash wants to test drones in the Mission. Some are pushing back. 
DoorDash plans to launch a drone testing site on Folsom Street, but the Teamsters Union Local No. 665 is pushing back. 

Claude Sonnet 4.5 shows significantly increased situational awareness when testing for alignment, here's a fascinating example from p. 59 of the system card (#anthropology #AIResearch

https://www.macrumors.com/2025/10/18/ios-26-1-to-ios-26-4-expected-features/

X is testing a new way of opening links without fully covering an X post, allowing users to see the Like, Repost, and other buttons, starting on iOS (Cheyenne MacDonald/Engadget)
https://www.engadget.com/x-is-testing-a-new-way-of…

X is testing a new way of opening links in posts to improve engagement
The new link popups will allow the reaction buttons to remain visible so users can interact with the post while reading.

Benchmarking PDF Accessibility Evaluation A Dataset and Framework for Assessing Automated and LLM-Based Approaches for Accessibility Testing
Anukriti Kumar, Tanushree Padath, Lucy Lu Wang
https://arxiv.org/abs/2509.18965

git #bisect-ing a kernel on real hardware sucks. I can at least be grateful I'm not testing on my main machine. The #firmware wait times are huge and the grub window so tiny you can't leave it to it. I need to apply patches at each stage that touch pgtable so almost everything needs to be re…

With this being the first Raspberry Pi 5 of the cluster, and since those are notorious for running hotter than the Pi 4. Which I did extensive testing on last year: https://blog.wyrihaximus.net/2024/12/building-a-kubernetes-homelab-w…

Zu Testzwecken läuft PixelFed jetzt erst einmal auf einem lokalen Rechner auf der Instanz https://pixelfed.pixelgalaxy.net
Sobald ein geeigneter VPS verfügbar ist, werde ich die Instanz auf der Hauptdomain installieren. Der Installationsprozess mit YunoHost war relativ einfach. Der Teufel liegt aber wie…

PixelGalaxy.NET - Testing
Testserver for later instance PixelGalaxy.NET

Crosslisted article(s) found for cs.ET. https://arxiv.org/list/cs.ET/new
[1/1]:
- An Empirical Study of Testing Practices in Open Source AI Agent Frameworks and Agentic Applications
Hasan, Li, Fallahzadeh, Rajbahadur, Adams, Hassan

Do you have a device with an AMD Neural Processing Unit (NPU) that you actively use? I could use your help testing support for monitoring AMD NPUs in Resources!
I'd really appreciate it if you'd check out the amdxdna-support branch of Resources' GitHub repository and check if NPU utilization correctly shows up in Resources.
You can find more info in this GitHub issue:

Fray Detects #Concurrency Issues in #JVM Languages
https://www.infoq.com/news/2025/12/fray-de

PSA about food labeling in the US
We have a gluten detection service dog because many things that should be gluten free/say they’re gluten free are not actually gluten free.
Stuff gets contaminated when growing (e.g. next to wheat field), by shared equipment, in factories, from packaging, during transport and in-store.
Every US consumer should know:
1. The list of ingredients on food isn't exhaustive
2. Allergen labeling:
a) limited to just some allergens
b) manufacturers don't actually have to test
c) "certified" foods are tested—but not continuously
d) testing only works with enough contamination
Some certifications may require batch-testing, but usually they don't.
A "certified gluten free" product may e.g. contain oats which sometimes are contaminated with gluten—but as not every batch is tested it's impossible to know unless you test yourself (hence the service dog).
Even if the product is properly batch-tested, you might get a part of the product that has the allergen in it, whereas the tested part didn't.
Or the threshold was too low (our dog can detect gluten better than any available lab testing equipment; yes, dogs are amazing).
Food products also contain ingredients that do not have to be included on the label when they're "incidental" (included in an another ingredient) or if they're considered part of the manufacturing process but not of the final product (e.g. various coatings on factory equipment).
Don't need to list flavors or specific spices either. ¯\_(ツ)_/¯
As for allergens, only those responsible for ~90% of food allergies* have to be specifically declared, and they're not tested for as it's simply based on the ingredients list.
Good luck if you have other allergies.
*milk, egg, egg, fish, Crustacean shellfish, tree nuts, wheat, peanuts, soybeans

Just deleted a bunch of testing boilerplate in this repo after @… added `import.meta.resolve` support 🏆
https://github.com/vitest-dev/vitest/issues/6953

Merde j'avais raté la sortie de servo 0.0.2 https://github.com/servo/servo/releases/tag/v0.0.2

Release v0.0.2 · servo/servo
v0.0.2 This release is equivalent to nightly 2025-11-14, with some additional manual testing of the release artifacts. For our first release, the version number received a lot of attention. We are ...

> The upshot is that, whereas it used to take upwards of 12 minutes to join Matrix HQ […] now down to about 30 seconds
In #matrix , room joins occur after you start to enjoy the loading spinner.
https://

Testing DAWN #きつねかわいい
https://fedibird.com/@noppe/115417355954696723

noppe :dawn: (@noppe@fedibird.com)
添付: 1 枚の画像 AppStoreでDAWN for Mastodonがフィーチャーされています！ これからMastodonを始める人も、直感的にSNSが楽しめるDAWNをぜひお使いください！！ https://apps.apple.com/jp/app/dawn-for-mastodon-by-nightfox/id1668645019 #きつねかわいい

Discover the power of property-based testing in R with the #quickcheck package! Seamlessly integrates with #testthat and offers a variety of generators for atomic vectors, lists, and tibbles. Perfect for ensuring your code's reliability. Check it out:

💥 Bubble wrap bursts enable power-free acoustic testing
#sensors

Bubble wrap bursts enable power-free acoustic testing
Non-destructive testing allows engineers to evaluate the integrity of structures such as pipelines, tanks, bridges, and machinery without dismantling them. Conventional approaches rely on loudspeakers, lasers, or electric sparks. While effective, these systems can be difficult or dangerous to use in flammable or confined areas and require considerable power to function effectively.

we are the clapton, always anti-fascist, fuck the terf FA, LET THEM PLAY
#claptoncfc #ccfc #ftfa #fedifc

Maybe this is it: we just need to rebrand Solar as Fusion (alpha testing) 😂 @… https://mastodon.world/@davidho/115400254100213642

David Ho (@davidho@mastodon.world)
Attached: 1 image At #NewScientistLive, we’re bombarded by ads that say: the climate crisis is terrible; fusion is the answer. Yes, fusion is the answer, but it’s not happening here on Earth but 8.33 minutes away.

PSA for users of #Fedora Linux 41 and 42:
The jump from #Linux 6.16.y to 6.17.y will soon happen in updates-testing; a few days later the new #kernel series will become a regular update.

Oh goody, a LinkedIn connection request with a message!
Let’s break this down:
> …vc backed…
Focused on quarterly profits / RoI instead of outcomes.
> …invite-only…
NDAs and other gag agreements.
> …openAI browser…
Chromium that begs authors for ARIA to parse content.
> …seeing promising results…
Which aren’t genuine results.
> …goal is 100% WCAG testing…
Ah, snake oil.
> …with high capture rates.
Sales…

Baidu and Swiss Post's PostBus plan to launch Baidu's Apollo Go autonomous vehicle service in Switzerland, testing in December ahead of rollout by Q1 2027 (Reuters)
https://www.reuters.com/technology/baidu-expands-robotaxi-push-swit…

Donation to your mastodon server will be cheaper https://techhub.social/@Techmeme/115737372852805627

"Regulators overlooking toxic Pfas found around Lancashire chemicals plant"
#UK #UnitedKingdom #PFAS #ForeverChemicals

Meta is testing limiting Facebook professional accounts and Pages to posting just two links per month, unless they subscribe to $14.99 /month Meta Verified (Ivan Mehta/TechCrunch)
https://techcrunch.com/2025/12/17/facebook-is-te…

Facebook is testing a link posting limit for professional accounts and pages | TechCrunch
Meta said that the test impacts professional accounts and pages on Facebook.

While testing some new misp-modules, such as the OpenAPI interface, I discovered a strange behavior in Firefox when trying to reach TCP port 6666, which is the default port used by misp-modules.
It seems Firefox blocks access to a predefined list of TCP ports, and this has been in place for quite some time, as you can see in the commit log.
If you want to override the blocked port list, there is an obscure setting called network.security.ports.banned.override.
…

Folks using #PhanpySocialDev , there are 2 new features that need a bit of testing:
1. QR code for profiles and shortcuts settings - includes scanner (camera) too
2. Import/export accounts - the export *excludes* access tokens, so need to login again after import
They’re quite hidden, so just a heads-up 🙇‍♂️

RE: https://mastodon.social/@firefoxwebdevs/115740501470592801
I agree in theory that "assume good faith“ is the right way to go and a very pragmatic maxim for communities, but Mozilla has been testing that assumption for quite a while now.

Testing some COB LEDs today…

As the system comes up, the component builders will from time to time appear,
bearing hot new versions of their pieces -- faster, smaller, more complete,
or putatively less buggy. The replacement of a working component by a new
version requires the same systematic testing procedure that adding a new
component does, although it should require less time, for more complete and
efficient test cases will usually be available.
-- Frederick Brooks Jr., "The Myt…

" @… noted that authoritarian regimes fear mass organizing and peaceful protest because they reveal a regime’s unpopularity and show that it is losing its grip on power."
"Much as tossing chests of tea into Boston Harbor did about 250 years ago." - @…

October 16, 2025
Yesterday the Trump administration announced it would pay furloughed troops by using funds Congress appropriated for research, development, testing, and evaluation (RDTE) for fiscal year 2026.

Sources: Meta plans to test an AI-powered personalized daily briefing, designed to compete with ChatGPT's Pulse, with some Facebook users in NYC and SF (Naomi Nix/Washington Post)
https://www.washingtonpost.com/technology/2025/11/21/meta-ai-powered-daily…

Meta is building an AI-powered morning brief in push to compete with ChatGPT
Meta is testing a new product that would give Facebook users a personalized, AI-powered briefing of information relevant to their daily lives.

from my link log —
Mutation testing for librsvg with cargo-mutants.
https://viruta.org/mutation-testing-librsvg.html
saved 2025-12-03 https://…

I’m somewhat exhausted to announce attrs 25.4.0!
The main reason for this release (and why it's published today) is that it ships the first pieces of work for Python 3.14 and PEP 749. There will be more work required and there's going to be a lot more churn once everyone starts testing 3.14 earnestly. We hope to receive more feedback before spending more time on this. #Python

Release 25.4.0 · python-attrs/attrs
Highlights The main reason for this release (and why it's published today) is that it ships the first pieces of work for Python 3.14 and PEP 749. There will be more work required and there's going ...

UK Unveils Roadmap to Phase Out Animal Testing, Commits £75M to Develop Alternative Methods https://vegconomist.com/science/uk-unveils-roadmap-phase-out-animal-testing-commiting-develop-alternative-methods/

UK Unveils Roadmap to Phase Out Animal Testing, Commits £75M to Develop Alternative Methods - vegconomist - the vegan business magazine
The UK government has unveiled a comprehensive roadmap that will allow animal testing to be phased out more quickly. Where possible, the strategy will empower

https://cyberscoop.com/bugcrowd-mayhem-security-acquistion-ai-security-testing/
Omg so happy for the great David Brumley!

sometimes it's fun making a little #ai #llm battle 😏 💪
i have no idea what these fancy benchmarks are testing, as a #php developer, i'm (most of the time) interested in the basics - so even a …

From Translink
New bus stop lights roll out to improve safety and visibility
TransLink testing solar-powered lights at 14 bus stops
TransLink is improving safety and comfort with new solar-powered lights at bus stops.

The project will test lights at 14 locations where customers or staff have identified a need for improved lighting. The initiative will improve visibility for both Bus Operators and waiting passengers.

Oh great, new issue spam attack on @…. People really have too much time on their hands. This time posting random images from 4chan, etc., as well as AI-generated bogus issues.
Sorry you guys have to deal with such crap.

I’m currently working on a software designed more than a decade ago. It offers a plugin architecture: you can develop a plugin whose lifecycle is handled by the software. The tough part, though, is how you access the platform capabilities: via static methods on singletons.
How did I #test my code?

Help test #deltachat calls
https://support.delta.chat/t/help-testing-upcoming-delta-chat-release-with-calls/4220

git #bisect-ing a kernel on real hardware sucks. I can at least be grateful I'm not testing on my main machine. The #firmware wait times are huge and the grub window so tiny you can't leave it to it. I need to apply patches at each stage that touch pgtable so almost everything needs to be re…

Meta is testing limiting Facebook professional accounts and Pages to posting just two links per month, unless they subscribe to $14.99 /month Meta Verified (Ivan Mehta/TechCrunch)
https://techcrunch.com/2025/12/17/facebook-is-te…

Facebook is testing a link posting limit for professional accounts and pages | TechCrunch
Meta said that the test impacts professional accounts and pages on Facebook.

I regularly warn that content on Forbes is pay-for-play.
Here an overlay vendor shares prompts to feed an LLM for testing code, demonstrating LLMs on their own can’t do it without extensive coaching _and_ that coaching needs to be correct (the examples have issues):
https://www.

AI Prompts For Web Accessibility Testing
Accessibility should never be treated as just a compliance checkbox; it should be about creating digital experiences that work for everyone.

LED Lighting: Mini Reviews - Real-world testing! - https://m.earth.org.uk/LED-lighting.html

LED Lighting: Mini Reviews
Real-world testing!

{testthat} is great for automatic testing. Here are some tricks for the heavy user: #rstats

😰 Miami Is Testing a Self-Driving Police Car That Can Launch Drones
https://www.thedrive.com/news/miami-is-testing-a-self-driving-police-car-that-can-launch-drones

Miami Is Testing a Self-Driving Police Car That Can Launch Drones
Meet the new surveillance state trooper!

With @… you gotta daily-check those repo's. Updated KDE liveslak, updated window manager goodies, updated testing (minimal KDE, with homebrew, flatpak and distrobox): on a roll!
#slackware

If I'm gonna do some "pen testing" it's gonna be me sitting at my desk with a stack of nice paper, some music playing, and no computers involved.

Just updated Node Pebble to support latest release version of Let’s Encrypt’s Pebble testing server.
#LetsEncrypt…

node-pebble
A Node.js wrapper for Let’s Encrypt’s Pebble (a small RFC 8555 ACME test server not suited for a production certificate authority)

PSA for users that regularly test #Fedora Beta as well as proposed updates once the new version was released:
Do not enable updates-testing[1] by modifying /etc/yum.repos.d/fedora-updates-testing.repo; instead do it like this:
$ sudo dnf config-manager setopt updates-testing.enabled=true
Otherwise updates-testing will be disabled shortly before the release of a new version (t…

The LAX Automated People Mover (APM) is a new, free, electric train system that connects airport terminals, parking, and public transportation.
After delays, it is expected to open for public use in January 2026,
with final testing concluding in June 2026.
The 2.25-mile elevated guideway has six stations, and trains will run 24/7, arriving every two minutes during peak hours

Agentic Property-Based Testing: Finding Bugs Across the Python Ecosystem
Muhammad Maaz, Liam DeVoe, Zac Hatfield-Dodds, Nicholas Carlini
https://arxiv.org/abs/2510.09907 https:/…

Agentic Property-Based Testing: Finding Bugs Across the Python Ecosystem
Property-based testing (PBT) is a lightweight formal method, typically implemented as a randomized testing framework. Users specify the input domain for their test using combinators supplied by the PBT framework, and the expected properties or invariants as a unit-test function. The framework then searches for a counterexample, e.g. by generating inputs and calling the test function. In this work, we demonstrate an LLM-based agent which analyzes Python modules, infers function-specific and cros…

Agentio, which uses AI to help brands automate and scale campaigns with YouTube creators, raised a $40M Series B led by Forerunner at a $340M valuation (Ivan Mehta/TechCrunch)
https://techcrunch.com/2025/11/18/agen

Agentio secures $40 million from Forerunner as it scales its creator marketplace beyond YouTube | TechCrunch
Agentio, which matches creators with brands, is expanding beyond YouTube and testing Meta platforms to facilitate sponsored content.

Function Health, a health tracking tech company, raised a $298M Series B led by Redpoint Ventures at a $2.5B valuation, bringing its total funding to $350M (Kate Park/TechCrunch)
https://techcrunch.com/2025/11/19/funct…

Function Health raises $298M Series B at $2.5B valuation | TechCrunch
Function, which offers blood testing and medical imaging services, recently raised $298 million in a Series B round led by Redpoint Ventures at a valuation of $2.5 billion.

LLMs are All You Need? Improving Fuzz Testing for MOJO with Large Language Models
Linghan Huang, Peizhou Zhao, Huaming Chen
https://arxiv.org/abs/2510.10179 https://

LLMs are All You Need? Improving Fuzz Testing for MOJO with Large Language Models
The rapid development of large language models (LLMs) has revolutionized software testing, particularly fuzz testing, by automating the generation of diverse and effective test inputs. This advancement holds great promise for improving software reliability. Meanwhile, the introduction of MOJO, a high-performance AI programming language blending Python's usability with the efficiency of C and C++, presents new opportunities to enhance AI model scalability and programmability. However, as a new l…

Reddit starts a limited test of verified profiles, an opt-in feature that places a gray checkmark beside the usernames of notable people or businesses (Amanda Silberling/TechCrunch)
https://techcrunch.com/2025/12/10/reddit-is-testing-verification-badges/

Reddit is testing verification badges | TechCrunch
Verified users will actually be notable figures, not just people who paid for the checkmark.

Google plans to release Gemini 3 Deep Think to Google AI Ultra subscribers in the coming weeks, once it passes further rounds of safety testing (Russell Brandom/TechCrunch)
https://techcrunch.com/2025/11/18/google-launches-gemini-3-…

Google launches Gemini 3 with new coding app and record benchmark scores | TechCrunch
Google released Gemini 3, its latest and most advanced foundation model, to be made immediately available through the Gemini app and AI search interface.

Agentic RAG for Software Testing with Hybrid Vector-Graph and Multi-Agent Orchestration
Mohanakrishnan Hariharan, Satish Arvapalli, Seshu Barma, Evangeline Sheela
https://arxiv.org/abs/2510.10824

Agentic RAG for Software Testing with Hybrid Vector-Graph and Multi-Agent Orchestration
We present an approach to software testing automation using Agentic Retrieval-Augmented Generation (RAG) systems for Quality Engineering (QE) artifact creation. We combine autonomous AI agents with hybrid vector-graph knowledge systems to automate test plan, case, and QE metric generation. Our approach addresses traditional software testing limitations by leveraging LLMs such as Gemini and Mistral, multi-agent orchestration, and enhanced contextualization. The system achieves remarkable accurac…

Windows 11 Copilot AI hands-on: despite Microsoft advertising the feature as able to "understand you", it delivers inconsistent and often incorrect responses (Antonio G. Di Benedetto/The Verge)
https://www.theverge.com/report/822443/mic

Talking to Windows’ Copilot AI makes a computer feel incompetent
Microsoft is advertising its Windows Copilot AI as “the computer you can talk to.” How does that hold up to testing, and how does it track with CEO Satya Nadella’s ambitions?

Amazon's Twitch begins testing livestream shopping ads allowing users to buy products "in real time", starting with e.l.f. Cosmetics (Peter Adams/Marketing Dive)
https://www.marketingdive.com/news/elf-cosmetics-first-test-twitc…

E.l.f. Cosmetics first to test Twitch’s new livestream shopping ads
A new integration powered by Amazon Ads’ retail media data is meant to drive “more measurable business outcomes” for E.l.f. on the platform. 

Agentio, which uses AI to help brands automate and scale campaigns with YouTube creators, raised a $40M Series B led by Forerunner at a $340M valuation (Ivan Mehta/TechCrunch)
https://techcrunch.com/2025/11/18/agen

Agentio secures $40 million from Forerunner as it scales its creator marketplace beyond YouTube | TechCrunch
Agentio, which matches creators with brands, is expanding beyond YouTube and testing Meta platforms to facilitate sponsored content.

Google tests AI-powered overviews on some publications' Google News pages; publishers like Der Spiegel, El País, and WaPo in commercial partnerships get paid (Aisha Malik/TechCrunch)
https://techcrunch.com/2025/12/10/goog

Google is testing AI-powered article overviews on select publications’ Google News pages | TechCrunch
By adding AI-powered article overviews, Google says users will get more context before they click through to read an article.

Shares of Tesla closed at a 2025 high on Monday after the company confirmed it is testing driverless vehicles in Austin without a human safety operator (Lora Kolodny/CNBC)
https://www.cnbc.com/2025/12/15/tesla-tests-driverless-cars-in-austin…

Tesla stock closes at 2025 high after Musk confirms driverless Robotaxi tests underway in Austin
Tesla CEO Elon Musk said in a social media post that his company is testing driverless cars in Austin without a safety driver in the vehicle.

WhatsApp is testing imposing per-month limits on how many messages individual users and businesses can send to unknown people without getting a response (Ivan Mehta/TechCrunch)
https://techcrunch.com/2025/10/17/what

WhatsApp will curb the number of messages people and businesses can send without a response | TechCrunch
The company said the controls are designed to be effective against people and businesses that blast messages and spam people.

Reddit starts a limited test of verified profiles, an opt-in feature that places a gray checkmark beside the usernames of notable people or businesses (Amanda Silberling/TechCrunch)
https://techcrunch.com/2025/12/10/reddit-is-testing-verification-badges/

Reddit is testing verification badges | TechCrunch
Verified users will actually be notable figures, not just people who paid for the checkmark.

Automaker Stellantis and Pony.ai sign a non-binding agreement to develop robotaxis for deployment in Europe, with plans to start testing in the coming months (Rebecca Bellan/TechCrunch)
https://techcrunch.com/2025/10/17/stellantis-teams-up-w…

Stellantis teams up with Pony AI to develop robotaxis in Europe  | TechCrunch
The collaboration will integrate Pony’s self-driving software into Stellantis’ electric medium-size van platform, which is built with advanced sensors to support autonomous vehicles.

Amazon's Twitch begins testing livestream shopping ads allowing users to buy products "in real time", starting with e.l.f. Cosmetics (Peter Adams/Marketing Dive)
https://www.marketingdive.com/news/elf-cosmetics-first-test-twitc…

E.l.f. Cosmetics first to test Twitch’s new livestream shopping ads
A new integration powered by Amazon Ads’ retail media data is meant to drive “more measurable business outcomes” for E.l.f. on the platform. 

Sources: Apple is in preliminary talks with India's CG Semi, which offers chip assembly and testing services, to assemble and package chips for the iPhone (Dia Rekhi/The Economic Times)
https://economictimes.indiatimes.com/t

Domestic firms may chip in for the iPhone
Exploratory conversations have taken place between Murugappa Group-owned CG Semi, which is building an outsourced semiconductor assembly and test (OSAT) facility in Sanand, Gujarat. It’s the first time Apple is evaluating the prospect of having certain chips assembled and packaged in India.

DoorDash launches Zesty, a standalone AI-powered social app for users to find nearby restaurants, in public testing in San Francisco and New York (Natalie Lung/Bloomberg)
https://www.bloomberg.com/news/articles/2025-12-…

Google tests AI-powered overviews on some publications' Google News pages; publishers like Der Spiegel, El País, and WaPo in commercial partnerships get paid (Aisha Malik/TechCrunch)
https://techcrunch.com/2025/12/10/goog

Google is testing AI-powered article overviews on select publications’ Google News pages | TechCrunch
By adding AI-powered article overviews, Google says users will get more context before they click through to read an article.

Amazon is testing an AI tool called Kindle Translate that automatically translates books into other languages, for authors that self-publish on the platform (Lawrence Bonk/Engadget)
https://www.engadget.com/ai/amazon-is-test

Amazon is testing an AI tool that automatically translates books into other languages
Amazon is introducing an AI tool that will automatically translate books into other languages. This should be useful for authors who self publish.

ROG Xbox Ally X review: great performance and grips, silent cooling, good 120Hz display, and USB-C, but Windows optimization isn't great and poor battery life (Rebecca Spear/Windows Central)
https://www.windowscentral.com/hardware/asus/asus-rog-xbox-ally-x-review…

Our review proves Xbox Ally X sets the bar in handheld PCs
We spent dozens of hours playing games and testing the performance on the Xbox Ally X. It's an impressive device worth paying attention to. Our review.